the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 29 Jun 2022
Aerosol data assimilation with aqueous chemistry in WRF-Chem/WRFDA V4.3.1
Abstract. This article introduces a new chemistry option in the Weather Research and Forecasting model data assimilation (WRFDA) system coupled with the WRF-Chem model (Version 4.3.1) to activate aqueous chemistry (AQCHEM) for the assimilation of surface concentrations of particulate matter (PM) along with atmospheric observations. The gas-phase mechanism used is the Regional Atmospheric Chemistry Mechanism (RACM), the inorganic aerosols are treated with the Modal Aerosol Dynamics Model for Europe (MADE), and secondary organic aerosol (SOA) production is parameterized based on the Volatility Basis Set (VBS) approach. The "RACM-MADE-VBS-AQCHEM" scheme used in the weakly coupled data assimilation and forecast system facilitates aerosol-cloud-radiation-precipitation interactions through analysis and forecast cycling, accounting for both direct and indirect aerosol effects in the short-term air quality prediction. The new implementation in the three-dimensional variational data assimilation (3D-Var) system was tested with the assimilation of PM2.5 and PM10 concentrations on the ground over the East Asian region through month-long cycling for Spring 2019. It is demonstrated that the inclusion of aerosol species in the aqueous (or cloud-borne) phase in both analysis and forecast reproduces aerosol wet removal processes in association with the development of clouds, systematically changing the atmospheric composition. The new option with aqueous chemistry in WRFDA is beneficial in air quality forecasting in cloudy conditions, while the simulations without aqueous chemistry overestimate surface PM10 (PM2.5) by a factor of 10 (3).
This preprint has been withdrawn.
-
Withdrawal notice
This preprint has been withdrawn.
-
Preprint
(2639 KB)
Interactive discussion
Status: closed
- RC1: 'Comment on egusphere-2022-371', Anonymous Referee #1, 29 Jul 2022
-
RC2: 'Comment on egusphere-2022-371', Anonymous Referee #2, 05 Sep 2022
The paper introduces a chemical data assimilation (DA) tool that is used with one of the full gas-aerosol chemistry schemes (RACM-MADE-VBS) included in the WRF-Chem model. This version of the gas/aerosol chemistry scheme also includes an aqueous chemistry parameterization to simulate cloud-borne aerosol species. This parameterization in the WRF-Chem enables simulating aerosol direct and indirect feedbacks. The chemical DA is a valuable technique to improve atmospheric composition simulations. Its application to the coupled meteorology-chemistry model with an advanced gas, aerosol and aqueous chemistry scheme can provide users with a powerful modeling tool that can be applied to forecasting air quality and chemistry-weater interactions. The chemical DA system and its application to the KORUS-AQ campaign deserves publication in GMD. However, there are shortcomings of the paper and modeling study that need to addressed first.
Major comments:
The paper is too long. There are some repetitions in the text. Some parts of the description of the chemical mechanism and model are already available in user’s guides and other papers (e.g. the paragraphs about WPS and WRF).
The introduction discusses the challenges associated with chemical DA. One of major limiting factor isn’t mentioned here is the very limited aerosol composition measurements. For example, the continuous and frequent (hourly) measurements of inorganic and organic aerosol species are very limited around the world.
The impact of clouds on the satellite AOD retrievals is discussed quite a bit. But only the ground-based in-situ PM observations are assimilated in the study, which aren’t susceptible to the cloudiness.
The author emphasizes the improvement of the simulations when the aqueous chemistry is used in conjunction with the gas and aerosol chemistry. The author explains this with the fact that the aqueous chemistry enables the aerosol wet removal processes in the model. It’s reported that PM10 is overestimated by a factor of 10 when the aqueous chemistry isn’t used. The wet removal processes of the aerosol species in the RACM-MADE-SOA scheme (without aqueous chemistry) in WRF-Chem are fully implemented in the standard version of the model, for both resolved and sub-grid precipitation. The author reports that “wet scavenging processes can be all simulated only when aerosols in cloud water are defined through aqueous chemistry, DA without aqueous chemistry treated all the aerosols as interstitial (e.g., suspended in the air) even when precipitation occurred, leading to a significant overestimation of surface PM concentrations.” It’s possible that the author didn’t take advantage of all the capabilities of the WRF-Chem model to simulate the removal processes of the gas and aerosol species in the base model configuration, thus overestimating the aerosol concentrations in the model simulations conducted without aqueous chemistry.
It should be noted that the use of the aqueous chemistry can provide such advantages as simulating cloud-phase sulfate production, aerosol-cloud interactions and so on. However, it’s wrong to claim that the WRF-Chem model lacks wet removal, unless aqueous chemistry is used. This would also imply that the previous WRF-Chem studies, where aqueous chemistry wasn’t used were wrong and produced inaccurate simulations.
Along these lines, the following statement in the Conclusion is also misleading: “By introducing aerosols in the aqueous (or cloud water) phase in WRFDA, the regional cycling system could represent wet scavenging in resolvable-scale cloud microphysics and convective transport.”
The presented WRF-Chem model configuration includes cumulus parameterization. However, the aqueous chemistry for the sub-grid clouds wasn’t included in the model. This issue needs to be discussed in the text.
Figure 2: Are the differences between the sensitivity simulations caused by the wet removal mostly? If this is the case, then I suggest conducting these simulations with the same chemistry option (108) with and without wet removal.
Line 205: this statement is confusing. The aerosol number concentrations have to be adjusted along with the mass concentrations in the model.
Although the inclusion of the aerosol-radiation and aerosol-cloud interactions in the model are highlighted in the text, the author doesn’t show any evaluation of the meteorological simulations. Thus it isn’t clear if using a computationally expensive chemistry scheme in WRF-Chem provides any advantages compared to other less complex schemes available in the model.
Minor comments:
Line 155: the composition of the unspeciated PM2.5 depends on the lumping of the species in the anthropogenic emission inventory, not just sea salt and dust.
Citation: https://doi.org/10.5194/egusphere-2022-371-RC2
Interactive discussion
Status: closed
- RC1: 'Comment on egusphere-2022-371', Anonymous Referee #1, 29 Jul 2022
-
RC2: 'Comment on egusphere-2022-371', Anonymous Referee #2, 05 Sep 2022
The paper introduces a chemical data assimilation (DA) tool that is used with one of the full gas-aerosol chemistry schemes (RACM-MADE-VBS) included in the WRF-Chem model. This version of the gas/aerosol chemistry scheme also includes an aqueous chemistry parameterization to simulate cloud-borne aerosol species. This parameterization in the WRF-Chem enables simulating aerosol direct and indirect feedbacks. The chemical DA is a valuable technique to improve atmospheric composition simulations. Its application to the coupled meteorology-chemistry model with an advanced gas, aerosol and aqueous chemistry scheme can provide users with a powerful modeling tool that can be applied to forecasting air quality and chemistry-weater interactions. The chemical DA system and its application to the KORUS-AQ campaign deserves publication in GMD. However, there are shortcomings of the paper and modeling study that need to addressed first.
Major comments:
The paper is too long. There are some repetitions in the text. Some parts of the description of the chemical mechanism and model are already available in user’s guides and other papers (e.g. the paragraphs about WPS and WRF).
The introduction discusses the challenges associated with chemical DA. One of major limiting factor isn’t mentioned here is the very limited aerosol composition measurements. For example, the continuous and frequent (hourly) measurements of inorganic and organic aerosol species are very limited around the world.
The impact of clouds on the satellite AOD retrievals is discussed quite a bit. But only the ground-based in-situ PM observations are assimilated in the study, which aren’t susceptible to the cloudiness.
The author emphasizes the improvement of the simulations when the aqueous chemistry is used in conjunction with the gas and aerosol chemistry. The author explains this with the fact that the aqueous chemistry enables the aerosol wet removal processes in the model. It’s reported that PM10 is overestimated by a factor of 10 when the aqueous chemistry isn’t used. The wet removal processes of the aerosol species in the RACM-MADE-SOA scheme (without aqueous chemistry) in WRF-Chem are fully implemented in the standard version of the model, for both resolved and sub-grid precipitation. The author reports that “wet scavenging processes can be all simulated only when aerosols in cloud water are defined through aqueous chemistry, DA without aqueous chemistry treated all the aerosols as interstitial (e.g., suspended in the air) even when precipitation occurred, leading to a significant overestimation of surface PM concentrations.” It’s possible that the author didn’t take advantage of all the capabilities of the WRF-Chem model to simulate the removal processes of the gas and aerosol species in the base model configuration, thus overestimating the aerosol concentrations in the model simulations conducted without aqueous chemistry.
It should be noted that the use of the aqueous chemistry can provide such advantages as simulating cloud-phase sulfate production, aerosol-cloud interactions and so on. However, it’s wrong to claim that the WRF-Chem model lacks wet removal, unless aqueous chemistry is used. This would also imply that the previous WRF-Chem studies, where aqueous chemistry wasn’t used were wrong and produced inaccurate simulations.
Along these lines, the following statement in the Conclusion is also misleading: “By introducing aerosols in the aqueous (or cloud water) phase in WRFDA, the regional cycling system could represent wet scavenging in resolvable-scale cloud microphysics and convective transport.”
The presented WRF-Chem model configuration includes cumulus parameterization. However, the aqueous chemistry for the sub-grid clouds wasn’t included in the model. This issue needs to be discussed in the text.
Figure 2: Are the differences between the sensitivity simulations caused by the wet removal mostly? If this is the case, then I suggest conducting these simulations with the same chemistry option (108) with and without wet removal.
Line 205: this statement is confusing. The aerosol number concentrations have to be adjusted along with the mass concentrations in the model.
Although the inclusion of the aerosol-radiation and aerosol-cloud interactions in the model are highlighted in the text, the author doesn’t show any evaluation of the meteorological simulations. Thus it isn’t clear if using a computationally expensive chemistry scheme in WRF-Chem provides any advantages compared to other less complex schemes available in the model.
Minor comments:
Line 155: the composition of the unspeciated PM2.5 depends on the lumping of the species in the anthropogenic emission inventory, not just sea salt and dust.
Citation: https://doi.org/10.5194/egusphere-2022-371-RC2
Model code and software
WRF-Chem/WRFDA Soyoung Ha https://doi.org/10.5281/zenodo.6569325
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 434 | 254 | 41 | 729 | 31 | 27 |
- HTML: 434
- PDF: 254
- XML: 41
- Total: 729
- BibTeX: 31
- EndNote: 27
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1